In the past few years, considerable progress has been made in our understanding of the sequence and structure of the genes and gene products of influenza viruses. Both surface glycoproteins of the virus that are important in antigenic variation have been crystallized and the three-dimensional structure determined by X-ray crystallography. Despite these achievements, influenza viruses still continue to appear that can cause high mortality in animal species - most recently in chickens. Influenza in humans has been relatively mild over the past decade, causing only modest increases in morbidity and mortality, but the potential exists for a virus of high pathogenicity to occur in humans. There are clearly gaps in our knowledge of influenza virus and the specific aims are the following: 1) To define the antigenic areas on the neuraminidase of influenza virus, especially the N9 subtype that possesses both hemagglutinating and enzyme activity and is phenotypically similar to the HN protein of paramyxoviruses. 2) To define the molecular basis of the biological functions of the hemagglutinin (HA), including the antigenic sites, receptor binding site and fusion peptide and relate them to influenza virus variability. 3) To determine how antigenic drift occurs in nature in the presence of a polyclonal antibody population and whether it can reach an evolutionary endpoint. 4) To define the molecular changes that permit an influenza virus to become virulent. The outbreak of H5N2 influenza in chickens in Pennsylvania that "acquired" virulence offers a natural system for such studies. Recent advances in sequencing methods and monoclonal antibodies now permit analysis of the important outstanding biological questions such as the origin of influenza viruses, the molecular basis of virulence and how antigenic drift occurs in nature.